Related Experiment Video
Updated: Jul 15, 2025
08:08
Detection and Isolation of Cancer in Prostate Biopsies Using Stimulated Raman Histology and Artificial Intelligence
Published on: June 10, 2025
72
Artificial Intelligence in Digital Pathology for Bladder Cancer: Hype or Hope? A Systematic Review.
Farbod Khoraminia1, Saul Fuster2, Neel Kanwal2
1Department of Urology, Erasmus MC Cancer Institute, University Medical Center Rotterdam, 3015 GD Rotterdam, The Netherlands.
Cancers
|September 28, 2023
Summary
Computational pathology (CPATH) offers an objective approach to improve bladder cancer (BC) prognosis by analyzing image features. While promising, further development is needed for clinical application.
Area of Science:
- Digital pathology
- Oncology
- Artificial Intelligence in Medicine
Background:
- Subjective pathological evaluation hinders accurate bladder cancer (BC) diagnosis and prognosis prediction.
- Computational pathology (CPATH) presents an objective method to identify clinical outcome predictors.
Purpose of the Study:
- To systematically review and analyze studies utilizing CPATH for bladder cancer (BC).
- To assess the potential of CPATH in improving BC diagnosis and prognosis prediction.
Main Methods:
- Systematic review of 2285 identified studies, analyzing 33 relevant CPATH studies in BC.
- Analysis of image features within regions of interest to differentiate tissue types, grades, and stages.
- Evaluation of cell nuclear features (area, shape irregularity, roundness) as prognostic markers.
Main Results:
- CPATH accurately distinguished normal vs. tumor tissue, tumor grade/stage, and tissue types (>80% accuracy).
- Nuclear features like area, shape irregularity, and roundness were strong predictors of recurrence and survival.
- CPATH identified molecular subtypes and improved recurrence/survival prediction when combined with clinicopathological data.
Conclusions:
- CPATH shows significant potential for enhancing bladder cancer diagnosis and prognosis.
- Further research requires more robust, interpretable models and larger, clinically representative datasets to ensure reliability and clinical applicability.
- Addressing AI's "black box" challenge is crucial for widespread clinical adoption of CPATH in oncology.
